Copula regression and robustness

This thesis aims to study the application of copula-based regression method and its robustness. In the first chapter, we will briefly review the fundamentals of copula theory with some important results. For the second chapter, we will combine some copula selection criteria and then simulate data sets from different kinds of copulas and apply the copula-based regression method on these data sets and demonstrate why it performs better than the non-parametric regression method. Subsequently, the third chapter will be the application on a real data set. Moreover, we will check the model's adequacy using another result from the goodness of fit for the copula models and apply the copula-based regression method to build a confidence interval of the prediction values. Finally, we will analyze and draw conclusion from these results and make some further research suggestions.

Embargo status: Restricted until August 2023. To request access, contact the author-- https://www.researchgate.net/publication/332736848_Robust_Predictive_Model_Using_Copulas